Phase III Trial of Immunotherapy with Recombinant Interleukin-2 (rIL-2) Versus Observation in Patients < 60 Years with Acute Myeloid Leukemia (AML) in First Remission (CR1): Preliminary Results from Cancer and Leukemia Group B (CALGB) 19808.

The CALGB evaluated the use of rIL-2 for immunotherapy of minimal residual disease in a phase III trial in patients with AML in first CR after completing all planned chemotherapy. The rationale supporting the use of rIL-2 in this setting includes its ability to effect antigen-independent cytotoxicity against AML blasts, its non-cross resistance with cytotoxic agents, and its ability to expand cytotoxic T and natural killer (NK) cells. Pts < 60 years with untreated non-M3 AML were eligible. 734 patients were enrolled. The first 302 patients were randomized between 2 induction regimens: Ara-C, Daunorubicin, and Etoposide (ADE) or ADEP with the P-glycoprotein modulator PSC-833 (Kolitz et al, ASH 2005). The remaining 432 patients received ADE induction. Post-remission therapy was based on cytogenetic risk factors: patients with Core Binding Factor (CBF) AML received 3 courses of High-Dose Ara-C (HiDAC), while all others were assigned to receive a 2 step autologous transplant (ASCT) regimen (Linker et al, Biol Blood Marrow Transpl 2000). Randomization between rIL-2 and observation was to occur no later than 120 days after day 1 of the last HiDAC cycle or day 0 of ASCT, as soon as the neutrophil count > 750/μl, platelets > 50,000/μl with bone marrow showing a leukemia-free state and trilineage maturation and recovery from non-hematological toxicity to < grade 2. The 90 day immunotherapy regimen consisted of low-dose rIL-2 sequences for expanding cytotoxic effector cells and brief, higher dose bolus treatments aimed at activating them. rIL-2 was given SC at 1 x 10⁶ IU/m² on days 1–14, 19–28, 33–42, 47–56, 61–70 and 75–90, and 12–15 x 10⁶ IU/m² on days 15–17, 29–31, 43–45, 57–59 and 71–73. CR was achieved in 77% of evaluable patients. After HiDAC consolidation or ASCT, patient refusal, early relapse, and delayed blood count recovery accounted for nearly all failures to undergo randomization to IL-2 or observation. The distribution of patients with CBF and non-CBF AML was comparable between the randomized arms. The median follow-up time from the post-remission randomization date for the surviving patients is 29 months. By intention-to-treat, for the 214 randomized patients, the 3-year disease-free survival rate is 45% (95% CI: 35%, 56%) on the observation arm and 56% (47%, 67%) on the rIL-2 arm (p=0.11; logrank test); the 3-year overall survival rate is 61% (52%, 72%) for patients randomized to observation and 68% (58%, 79%) for the rIL-2 arm (p=.0.09). Twenty-nine of the 107 patients randomized to rIL-2 therapy either refused to receive rIL-2 or were unable to start because of unresolved toxicities; another 28 patients started treatment but failed to complete their 90-day course. Grade 4 toxicities were neutropenia (17%), thrombocytopenia (11%), febrile neutropenia (FN, 1%), increased bilirubin (1%) and hypocalcemia (1%). Grade 3 toxicities, observed in 10%–14% of patients, were hypotension, fatigue, dehydration and FN. We conclude that post-consolidation immunotherapy with 90 days of rIL-2 is tolerable but not well accepted by patients and/or physicians. Further follow-up and additional analyses are planned, correlating outcomes with clinical subsets, amount of rIL-2 therapy received, as well as measurements of ex vivo cytotoxicity mediated by patients’ effector cells against cryopreserved autologous AML blasts.